Leak detection sensing for liquid cooled servers

Abstract

The invention relates to leak detection sensing for liquid cooled servers. Cooling systems and methods for detecting coolant leaks are disclosed. A processor of the server is attached to the cold plate. An internal liquid coolant is contained in a coolant reservoir and circulates through the cold plate via a coolant distribution manifold. Coolant leaks are detected based at least on a level of the internal liquid coolant in a tapered chamber of the coolant distribution manifold.

Description

[0001] Cross-reference of related applications

[0002] This application is a partial continuation-in-process of U.S. Application No. 18 / 906,041, filed on October 3, 2024, which is incorporated herein by reference in its entirety. Technical Field

[0003] This disclosure relates to liquid cooling of electronic devices. Background Technology

[0004] Electronic devices generate significant heat during operation, necessitating cooling systems to prevent overheating. Basic cooling systems use fans to blow air across the heat-generating components of the electronic device. More advanced cooling methods include liquid cooling, where heat is transferred from the electronic device to a liquid coolant, which then carries the heat to an external heat exchanger for dissipation. In immersion cooling, the entire electronic device is submerged in a bath of liquid coolant. In contrast, in direct liquid cooling, the components of the electronic device are attached to a cold plate through which liquid coolant circulates. While direct liquid cooling is advantageous in some applications, it carries the risk of coolant leakage, which could damage the electronic device. Summary of the Invention

[0005] In one embodiment, the server's processor is attached to a cold plate. Internal liquid coolant is contained in a coolant reservoir and circulates through the cold plate via a coolant distribution manifold. Coolant leaks are detected based at least on the level of the internal liquid coolant in the tapered chamber of the coolant distribution manifold. Attached Figure Description

[0006] Figure 1 A block diagram illustrating a server cooling system according to an embodiment of the present invention.

[0007] Figure 2 An isometric view of a server rack according to an embodiment of the present invention is shown.

[0008] Figure 3 A schematic representation of a coolant distribution manifold according to an embodiment of the present invention is shown.

[0009] Figure 4 A coolant reservoir containing an internal liquid coolant is shown according to an embodiment of the present invention.

[0010] Figure 5 A schematic representation of direct liquid cooling of a processor according to an embodiment of the present invention is shown.

[0011] Figure 6 Demonstrating attachment according to an embodiment of the invention Figure 5 A top view of the processor's cold plate.

[0012] Figure 7 A flowchart illustrating a method for detecting leakage of internal liquid coolant in a liquid-cooled server according to an embodiment of the present invention.

[0013] Figure 8 A flowchart illustrating a method for detecting leakage of internal liquid coolant in a liquid-cooled server according to an embodiment of the present invention.

[0014] Figure 9 A schematic representation of a coolant distribution manifold according to another embodiment of the present invention is shown.

[0015] Figure 10 Demonstrating embodiments according to the present invention Figure 9 A schematic cross-sectional view of the liquid level sensing section of the coolant distribution manifold.

[0016] Figure 11 A flowchart illustrating a method for detecting leakage of internal liquid coolant in a liquid-cooled server according to another embodiment of the present invention.

[0017] Figure 12 A block diagram of a computer that can be used in conjunction with embodiments of the present invention is shown. Detailed Implementation

[0018] In this disclosure, numerous specific details, such as examples of systems, materials, components, structures, and methods, are provided to provide a thorough understanding of embodiments of the invention. However, those skilled in the art will recognize that the invention can be practiced without one or more of these specific details. In other instances, well-known details have not been shown or described to avoid obscuring aspects of the invention.

[0019] Figure 1 A block diagram of a server cooling system 100 according to an embodiment of the present invention is shown. The cooling system 100 provides liquid cooling to a plurality of servers 120. In one embodiment, the cooling system 100 includes a coolant distribution unit (CDU) 150, a coolant distribution manifold (CDM) 130, and a plurality of cold plates (e.g., see...). Figure 5 and 6 Cold-rolled steel plate 502).

[0020] In one embodiment, each of the servers 120 is a server computer (i.e., hardware) having one or more processors cooled by direct liquid cooling. Specifically, the processors or other high-power components of the server 120 are attached to a cold plate. An internal liquid coolant is circulated through internal channels of the cold plate. Heat from the processor is conducted to the cold plate and thus to the internal liquid coolant. A leak sensor 121 detects a leak when internal liquid coolant leaks into the server 120.

[0021] CDU 150 may include a pump 151, a coolant reservoir 152, a control processor 153, and a heat exchanger 154. The coolant reservoir 152 contains an internal liquid coolant, which is circulated in an auxiliary cooling loop 135 by the pump 151. The internal liquid coolant preferably has a low conductivity, for example, less than 5 μS / cm. In this way, damage to electronic components is minimized in the event of coolant leakage. The internal liquid coolant may include propylene glycol, water, and additives (e.g., corrosion inhibitors), which together produce a conductivity of less than 5 μS / cm. The weight percentage of specific additives and components in the internal liquid coolant depends on the specific cooling requirements.

[0022] exist Figure 1 In this example, an external liquid coolant (e.g., water) is supplied by cooling tower 170. The external liquid coolant circulates in the main cooling loop 171. Heat from the internal liquid coolant is transferred to the external liquid coolant via heat exchanger 154.

[0023] CDM 130 dispenses internal liquid coolant to server 120. CDM 130 includes inlet 132, outlet 134, and fitting 131. Fitting 131 connects to a piping system that delivers the internal liquid coolant to a cold plate attached to the processor of server 120. Figure 1 In this example, internal liquid coolant enters through inlet 132, circulates through fitting 131 across the cold plate of server 120, and exits through outlet 134 to flow back to coolant reservoir 152. The heated internal liquid coolant is cooled by external liquid coolant supplied by cooling tower 170 via heat exchanger 154.

[0024] In one embodiment, the status of cooling tower 170 is reported to control server 180 (see [link]). Figure 1 Line 106) enables the server management software 181 to monitor the flow rate of the external liquid coolant, the temperature of the external liquid coolant, the pressure in the main cooling loop 171, and other conditions that may affect the operation of the cooling system 100.

[0025] Leak sensor 121 detects leaks of internal liquid coolant in server 120. For example, leak sensor 121 may be located on a cold plate and send an alarm when triggered (e.g., when one or more droplets of internal liquid coolant come into contact with leak sensor 121). In one embodiment, a baseboard management controller (BMC) of server 120 monitors the status of leak sensor 121 in server 120 and reports the status of leak sensor 121 to control server 180 (see [link to documentation]). Figure 1 (Line 101).

[0026] The level sensor monitors the level of the internal liquid coolant in the CDM 130 (see...). Figure 3 Level sensor 302). At least one other level sensor (see Figure 4 Level sensors 352 and 353 monitor the level of internal liquid coolant in the coolant reservoir 152. These internal liquid coolant level sensors (also simply referred to as "coolant level sensors") are triggered to send an alarm when the internal liquid coolant level drops below a predetermined threshold level. The control processor 153 is electrically connected to the coolant level sensors in the CDM 130 (see [link to relevant documentation]). Figure 1 (see line 102) and the coolant level sensor in the coolant reservoir 152 (see line 102) and the coolant level sensor in the coolant reservoir 152. Figure 1 (line 103), so that the control processor 153 can detect the trigger when the coolant level sensor is triggered.

[0027] The control processor 153 may be a microcontroller, a central processing unit (CPU), or other processor. The control processor 153 has associated memory (not shown) storing instructions for performing the functionality of the control processor 153 described herein. In one embodiment, the control processor 153 is configured to report the status of the coolant level sensor (see arrow 104) to a control server 180 via a computer network. Typically, the status of the leak sensor 121 and the coolant level sensor indicates whether the sensor has been triggered.

[0028] In one embodiment, a control server 180 hosts server management software 181, which manages server 120 as part of a data center. The server management software 181 is configured to detect leaks in internal liquid coolant based on the status of a leak sensor 121 and a coolant level sensor. The server management software 181 is configured to perform or initiate intervention in response to the detection of an internal liquid coolant leak. The intervention depends on the severity of the coolant leak, based on the status of the leak sensor 121 and the coolant level sensor.

[0029] As an intervention, server management software 181 can perform a normal shutdown or an immediate shutdown of server 120. A normal shutdown allows the operating system of server 120 to properly shut down all running programs and services before the power to server 120 is cut off, thereby preventing data corruption or loss. In contrast, an immediate shutdown immediately cuts off the power to server 120. Server management software 181 can send a command to the BMC of server 120 to perform a normal shutdown of server 120.

[0030] Server 120 can be rack-mounted (see...) Figure 2 In rack 201. Power distribution unit (PDU) 140 can provide power to all servers 120 in the rack. Server management software 181 can send a command to power distribution unit 140 to immediately cut off the power to all servers 120 in the rack and thus shut them down immediately (see line 105).

[0031] Figure 2 An isometric view of a server rack 201 according to an embodiment of the present invention is shown. The rack 201 has features for accepting servers 120 ( Figure 2 (Multiple levels not shown in the text) Figure 2 The image shows CDU 150 and CDM 130. CDM 130 is vertically mounted in rack 201. For example, fitting 131 may be a quick-connect fitting. A piping system (not shown; e.g., hoses) connects fitting 131 to the cold plate of the corresponding server 120. Inlet 132 and outlet 134 of CDM 130 are located at the bottom of CDM 130 (see dashed box 203). Piping system 204 (e.g., hoses) connects CDM 130 to coolant reservoir 152 in CDU 150. See reference... Figure 3 As explained, the top of the CDM 130 (see dashed box 202) may have a level sensor for detecting the level of the internal liquid coolant in the CDM 130.

[0032] Figure 3A schematic representation of a CDM 130 according to an embodiment of the invention is shown. The CDM 130 includes fittings 303 and 304, a fitting 131, an inlet 132, an outlet 134, a level sensor 302, and a coolant observation window 301. Internal liquid coolant enters through the inlet 132, flows out from the fitting 131 located on the fitting 303 to circulate through the cold plate of the server 120, exits from the cold plate to enter the fitting 131 located on the fitting 304, and exits through the outlet 134. In one embodiment, fittings 303 and 304 are made of stainless steel. The coolant observation window 301 is made of a transparent material (e.g., glass or PVC sheet) compatible with the internal liquid coolant. The coolant observation window 301 advantageously allows a user to visually inspect the level of the internal liquid coolant in the CDM 130, which is vertically mounted in a server rack 201, with the level sensor 302 and the coolant observation window 301 positioned facing upwards.

[0033] The level sensor 302 serves as a coolant level sensor in the CDM 130. The level sensor 302 is triggered to send an alarm when the internal liquid coolant level drops below a predetermined threshold, said threshold being... Figure 3 In this example, the position of the level sensor 302 within the pipe fitting 303 is set. The level sensor 302 is electrically connected to the control processor 153 (see [link]). Figure 1 and 3 (line 102), thereby enabling the control processor 153 to detect the trigger when the level sensor 302 is triggered and thus notify the server management software 181. In one embodiment, the server management software 181 issues a warning when the level sensor 302 is triggered. The warning allows the user to be notified to visually inspect the internal liquid coolant level through the observation window 301 and to refill the internal liquid coolant if necessary.

[0034] It should be noted that seepage or other normal conditions occurring relative to fittings 303 and 304 can lead to gradual depletion of the internal liquid coolant, thereby triggering the level sensor 302. Therefore, triggering of the level sensor 302 usually indicates the need to refill the internal liquid coolant. However, if a leak occurs, the internal liquid coolant level can drop more rapidly, which is used in embodiments of the invention to verify the triggering of the leak sensor 121 to detect a leak.

[0035] Figure 4 A coolant reservoir 152 according to an embodiment of the present invention is shown. Figure 4In one example, the coolant reservoir 152 includes an observation window 351 that allows a user to visually inspect the level of the internal liquid coolant in the coolant reservoir 152 of the CDU 150. The coolant reservoir 152 further includes a primary level sensor 352 and a critical level sensor 353 for electrically monitoring the level of the internal liquid coolant. The primary level sensor 352 is triggered to send an alarm when the level of the internal liquid coolant drops below a predetermined threshold level. Figure 4 In this example, the position of the primary level sensor 352 within the coolant reservoir 152 is set. Similarly, the critical level sensor 353 is triggered to send an alarm when the level of the internal liquid coolant drops below a predetermined threshold level. Figure 4 In this example, the position of the critical level sensor 353 in the coolant reservoir 152 is set. The main level sensor 352 and the critical level sensor 353 are electrically connected to the control processor 153 (see [link]). Figure 1 Line 103; Figure 4 (Lines 103-1, 103-2), so that the control processor 153 can detect the trigger when the liquid level sensor is triggered.

[0036] exist Figure 4 In this example, the critical level sensor 353 is positioned much lower than the primary level sensor 352. For instance, in a coolant reservoir 152 containing 5 liters of internal coolant, triggering the primary level sensor 352 indicates a loss of 1 liter of internal coolant, while triggering the critical level sensor 353 indicates a loss of 4 liters of internal coolant. While triggering the primary level sensor 352 typically indicates that only refilling of the internal coolant is needed, triggering the critical level sensor 353 indicates a significant loss of internal coolant and therefore requires immediate intervention.

[0037] Figure 5A schematic representation of direct liquid cooling of a processor 503 in a server 120 according to an embodiment of the present invention is shown. The processor 503 may be a central processing unit (CPU), a graphics processing unit (GPU), or other high-power integrated circuit. The processor 503 is mounted on a circuit board 506 of the server 120. The circuit board 506 may be a printed circuit board (PCB) used as a motherboard. A cold plate 502 is attached to the processor 503. A piping system 505 (e.g., a hose) delivers internal liquid coolant to the cold plate 502 through a liquid port 504. A leak sensor 121 is positioned near the interface between the port 504 and the piping system 505 to detect a leak when internal liquid coolant leaks at the interface. The leak sensor 121 may be a resistive, capacitive, or other type of sensor that detects one or more droplets of internal liquid coolant falling onto the leak sensor 121. The status of the leak sensor 121 is transmitted, for example, by the BMC of the corresponding server 120 to the control server 180 (see [link]). Figure 1 Line 101; Figure 6 (line 101) so that the server management software 181 can receive a notification when the leakage sensor 121 is triggered.

[0038] Figure 6 Demonstrating embodiments according to the present invention Figure 5 A top view of the cold plate 502. Figure 6 In this example, the internal liquid coolant flows from CDM 130, enters the inlet port 504-1 of the cold plate 502 via piping system 505-1, circulates through the cold plate 502, exits through the outlet port 504-2 of the cold plate 502, and flows back to CDM 130 via piping system 505-2. A leak sensor 121 is installed on the cold plate 502 below piping systems 505-1 and 505-2. Figure 6 In this example, leak sensor 121 surrounds the interface between port 504 and piping system 505, where coolant leakage is most likely to occur. When internal liquid coolant leaks at the interface, one or more droplets of internal liquid coolant fall onto leak sensor 121 and trigger the leak sensor.

[0039] Figure 7 A flowchart illustrating a method 550 for detecting leakage of internal liquid coolant in a liquid-cooled server according to an embodiment of the present invention is provided. Method 550 can be implemented by server management software 181 in conjunction with the control processor 153 of a CDU 150 (shown in...). Figure 1 (The method is executed by the component in the middle). It is understood that method 550 can also be executed by other components without diminishing the advantages of the present invention.

[0040] exist Figure 7In this example, the status of the leak sensor 121 in server 120, the level sensor 302 in CDM 130, the main level sensor 352 in coolant reservoir 152, and the critical level sensor 353 in coolant reservoir 152 are monitored. The sensors may be in a triggered state or a normal (i.e., untriggered) state.

[0041] When the leakage sensor 121 of server 120 ( Figure 7 ,551) and the liquid level sensor 302 in CDM 130 ( Figure 7 ,552) is simultaneously in a triggered state ( Figure 7 When performing a logical AND operation (553), the server management software 181 initiates a normal shutdown of all servers 120 located in the same rack as server 120. Figure 7 ,554).

[0042] It should be noted that alarms from leak sensors typically indicate that internal liquid coolant or another liquid has come into contact with the leak sensor within the server. In conventional coolant leak detection systems, intervention is performed to shut down the server in response to an alarm received from a leak sensor. However, this alarm may be a false alarm, meaning it does not necessarily indicate that internal liquid coolant is leaking within the server. Specifically, moisture, electrical interference, or other irrelevant conditions can cause the leak sensor to trigger. In method 550, alarms from leak sensors are verified by checking alarms from the internal liquid coolant level sensor in the CDM 130. This method advantageously prevents unnecessary shutdowns due to false alarms, thereby avoiding computational time losses and potential data corruption.

[0043] Server 120 may take some time to complete a normal shutdown. As a safeguard, to prevent permanent damage to server 120 if the normal shutdown takes too long or cannot be completed for some reason, server management software 181 starts a shutdown timer (e.g., five minutes) when the normal shutdown is initiated. After the shutdown timer expires, server management software 181 disconnects the power supply to the rack housing server 120. Figure 7 (555) and immediately shut down server 120. In an embodiment where the BMC of server 120 monitors the status of the leakage sensor 121, immediately shutting down server 120 after a predetermined time advantageously allows for a trigger to be reported before the BMC is damaged due to leakage. It is understood that it is likely that a normal shutdown will complete before the shutdown timer expires and the rack power is cut off, shutting down server 120.

[0044] When the leakage sensor 121 of server 120 ( Figure 7 ,551), level sensor 302 in CDM 130 ( Figure 7 The main liquid level sensor 352 in the coolant reservoir 152 (552) and the main liquid level sensor 352 in the coolant reservoir 152 Figure 7 ,556) is simultaneously in a triggered state ( Figure 7 When performing a logical AND operation (557), the server management software 181 cuts off the power supply to the rack housing the server 120. Figure 7 ,555) and immediately shut down server 120.

[0045] To account for potential fluctuations in the level of the internal liquid coolant in the coolant reservoir 152, the server management software 181 may wait for two or more alarms from the main level sensor 352 before confirming that the main level sensor 352 has been triggered. For example, after receiving a signal from the control processor 153 that the main level sensor 352 has been triggered, the server management software 181 may poll the status of the main level sensor 352 at least once more within a predetermined time window or wait for the control processor 153 to indicate at least once more that the main level sensor 352 has been triggered to confirm that the main level sensor 352 has been triggered.

[0046] The internal liquid coolant in coolant reservoir 152 may gradually decrease during normal operation. However, triggering of leak sensor 121, level sensor 302 in CDM 130, and main level sensor 352 in coolant reservoir 152 of CDU 150 indicates a serious coolant leak. Therefore, in this situation, server 120 is immediately shut down instead of initiating a normal shutdown first.

[0047] When the critical liquid level sensor 353 in the coolant reservoir 152 is triggered ( Figure 7 When (558), the server management software 181 cuts off the power to the rack housing the server 120. Figure 7 The server 120 is immediately shut down because the critical level sensor 353 indicates a significant loss of internal liquid coolant. The server 120 is shut down immediately regardless of the status of the leak sensor 121, the level sensor 302 in the CDM 130, and the main level sensor 352 in the coolant reservoir 152.

[0048] Figure 8 A flowchart illustrating a method 600 for detecting leakage of internal liquid coolant in a liquid-cooled server according to an embodiment of the present invention is provided. Method 600 may be performed by a computer (e.g., a control server 180 running server management software 181). Method 600 is explained in the context of a single server. It will be understood that method 600 may be performed for multiple servers.

[0049] In step 601, the cold plate is attached to the server's processor.

[0050] In step 602, the internal liquid coolant is allowed to flow through the cold plate.

[0051] In step 603, leakage of internal liquid coolant in the server is monitored. In one embodiment, the leak sensor is triggered in response to the detection of one or more droplets of internal liquid coolant falling onto a leak sensor attached to a cold plate. The triggering of the leak sensor causes it to send a corresponding alarm.

[0052] In step 604, the level of the internal liquid coolant in the CDM is monitored. In one embodiment, the coolant distribution manifold is vertically mounted in the rack housing the server, and the level sensor in the CDM is triggered to send an alarm when the level of the internal liquid coolant in the CDM drops below a first threshold level.

[0053] In step 605, the level of the internal liquid coolant in the CDU reservoir (i.e., the coolant reservoir in the CDU) is monitored. In one embodiment, the internal liquid coolant is contained in the CDU reservoir and flows through the cold plate via a coolant distribution manifold. When the level of the internal liquid coolant in the CDU reservoir drops below a second threshold level, a level sensor in the CDU reservoir is triggered to send an alarm.

[0054] In step 606, in response to the detection of leakage of internal liquid coolant in the server and the detection that the liquid level of internal liquid coolant in the CDM has dropped below a first threshold level, a normal shutdown of the server is initiated.

[0055] In step 607, in response to the detection of a leak of internal liquid coolant in the server, the detection that the level of internal liquid coolant in the CDM has dropped below a first threshold level, and the detection that the level of internal liquid coolant in the CDU storage has dropped below a second threshold level, an immediate shutdown of the server is initiated. In one embodiment, the immediate shutdown of the server is performed by cutting off the power supply to the server.

[0056] In some cooling applications, it can be advantageous to sense coolant leaks at a single location within the cooling system, rather than at multiple locations. For example, detecting coolant leaks based on the internal liquid coolant level in the CDM reduces the number of sensors and the complexity of the cooling system, eliminating the need to rely on detecting coolant leaks in the server and / or the internal liquid coolant level in the CDU storage. More specifically, in the aforementioned example, one or more coolant level sensors in the CDM will enable the detection of coolant leaks throughout the entire server rack in which the CDM is installed.

[0057] Figure 9 A schematic representation of a CDM 130A according to an embodiment of the present invention is shown. CDM 130A is... Figure 3 A specific embodiment of the CDM130. The CDM 130A is the same as the CDM 130, except that a level sensing section 650 is added to the CDM 130A. As will be more clearly described below, the level sensing section 650 includes a tapered chamber in which the level of the internal coolant is monitored by one or more level sensors.

[0058] In one embodiment, an vent valve 670 is positioned at the top of the level sensing section 650. The vent valve 670 vents the level sensor in the converging chamber and the auxiliary cooling loop, ensuring that the detection of the internal coolant level in the CDM 130A is unaffected by any positive or negative pressure areas in the auxiliary cooling loop. More specifically, in the event of a coolant leak, the vent valve 670 allows trapped air to escape as the internal coolant level drops, preventing air from becoming trapped within the level sensing section 650 and the auxiliary cooling loop as a whole, which could otherwise interfere with accurate level sensing or cause pressure imbalances within the cooling system. By releasing air, the vent valve 670 ensures that the level sensor in the level sensing section 650 continues to accurately monitor the internal coolant level even when the internal coolant level drops due to a leak.

[0059] In one embodiment, the level sensing section 650 extends from the top of the main section of the CDM 130A, the main section being... Figure 9 In this example, fitting 303 is used. It should be noted that the level sensing section 650 may also extend from the top of fitting 304. The level sensing section 650 includes a coolant viewing window 651 made of a transparent material (e.g., glass or PVC sheet) compatible with the internal liquid coolant. The coolant viewing window 651 advantageously allows a user to visually inspect the level of the internal liquid coolant in the CDM130A, which is vertically mounted in a server rack, with the level sensing section 650 positioned at the top.

[0060] Figure 10 A schematic cross-sectional view of a level sensing section 650 according to an embodiment of the present invention is shown. The level sensing section 650 includes an interface portion 652 coupled to a main section of a CDM 130A, which in one embodiment may be a fitting 303 or a fitting 304. Figure 10In this example, the interface portion 652 is threaded to facilitate direct coupling to the currently deployed CDM main section. The interface portion 652 can also be an accessory to allow the level sensing section 650 to be coupled to the main section via a piping system (e.g., a hose). Alternatively, the level sensing section 650 can be integrated with the main section of the CDM 130A, for example, in a one-piece design that includes the level sensing section.

[0061] The level sensing section 650 may be made of the same material as the main section (e.g., stainless steel or other materials compatible with the internal liquid coolant). The level sensing section 650 has a tapered chamber 654, the volume of which decreases towards the top. In one embodiment, the chamber 654 forms a conical funnel, with the funnel opening facing the main section and the funnel tip facing the exhaust port of the exhaust valve 670. The tapered shape of the chamber 654 advantageously achieves enhanced coolant level detection sensitivity.

[0062] In one embodiment, the level sensing section 650 includes one or more level sensors 653 (i.e., 653-1, 653-2, 653-3) for monitoring the level of the internal liquid coolant in the chamber 654. The level sensors 653 are electrically connected (see...). Figure 10 The control processor 153 of the CDU 150 (lines 102-1, 102-2, 102-3) is connected to the server management software 181 to monitor and report the status of the level sensor 653.

[0063] Each level sensor 653 triggers an alarm when the internal liquid coolant drops below a threshold, which in one embodiment is set by the location of the specific level sensor 653 within the chamber 654. The level sensors 653 can be used to monitor the rate of coolant drop. The coolant drop rate (from a first threshold level at level sensor 653-1 to a second threshold level at level sensor 653-2 and then to a third threshold level at level sensor 653-3) can be calculated and compared with the threshold rate to detect coolant leakage within the cooling system. A coolant drop rate exceeding the threshold rate indicates a coolant leakage. The threshold rate depends on the details of the cooling system.

[0064] In some embodiments, a single level sensor 653 is used to monitor the level of internal liquid coolant in chamber 654. In these embodiments, internal liquid coolant is added to an auxiliary cooling loop until it reaches a predetermined level in chamber 654. The single level sensor 653 is then configured such that a decrease in internal liquid coolant from the predetermined level to a threshold level of the single level sensor 653 corresponds to coolant loss indicating a coolant leak. In some embodiments, the single level sensor 653 may be combined with other detection methods (e.g., a leak sensor within server 120) to prevent false alarms.

[0065] The vent valve 670 may include a body 672 and a float 673. The body 672 has a top vent 671 and a bottom vent 674. Air can move through vents 671 and 674 to enter or exit chamber 654. During normal operation, when the internal liquid coolant level in chamber 654 is sufficient, the float 673 keeps the vent valve 670 closed. However, in the event of a coolant leak, when the internal liquid coolant level in chamber 654 decreases, the float 673 lowers and causes the vent valve 670 to open, allowing air to enter from vent 671. Once the internal liquid coolant level rises again, the float 673 causes the vent valve 670 to return to its closed position to maintain a system seal and prevent coolant loss.

[0066] In one embodiment, coolant leaks are detected solely based on the level of the internal liquid coolant in chamber 654. In other words, coolant leak detection in chamber 654 does not necessarily require confirmation or verification from a sensor in another location. This advantageously reduces the number of sensors required for leak detection.

[0067] In one embodiment, server management software 181 may perform coolant leak detection and initiate or perform intervention based on the level of the internal liquid coolant in CDM 130A. Server management software 181 may wait for an initial stabilization period (e.g., 12 hours) to complete before initiating leak detection. After the stabilization period, server management software 181 may receive the status of level sensor 653 from control processor 153 of CDU 150 and calculate the rate of decrease of the internal liquid coolant in chamber 654. Server management software 181 may initiate intervention, for example, to initiate a normal shutdown of all servers 120 in the rack where CDM 130A is installed, when the coolant decrease rate allows for a normal shutdown. To prevent permanent damage to servers 120, intervention may include immediately cutting off power to all servers 120 in the rack when the coolant decrease rate far exceeds a threshold rate.

[0068] Figure 11 A flowchart illustrating a method 690 for detecting leakage of internal liquid coolant in a liquid-cooled server according to an embodiment of the present invention is provided. Method 690 can be implemented by server management software 181 in conjunction with the control processor 153 of a CDU 150 (shown in...). Figure 1 (The method is executed by the component in the middle). It is understood that method 690 can also be executed by other components without diminishing the advantages of the present invention.

[0069] In step 691, the cold plate is attached to the server's processor.

[0070] In step 692, the internal liquid coolant in the coolant reservoir is allowed to flow through the cold plate via the CDM.

[0071] In step 693, the level of the internal liquid coolant is monitored in a tapered chamber of the CDM. In one embodiment, the tapered chamber forms a conical funnel. The tapered chamber may be located in a level sensing section that is directly and detachably (e.g., by thread) connected to the main section of the CDM, connected to the main section of the CDM via a piping system, or integrated with the main section of the CDM.

[0072] In step 694, a leak in the internal liquid coolant is detected at least based on the level of the internal liquid coolant in the converging chamber of the CDM. In one embodiment, one or more level sensors are used to monitor the level of the internal liquid coolant. A coolant leak is detected when one or more of the level sensors are triggered at a rate indicating that the coolant drop rate exceeds a threshold rate. The coolant drop rate can be calculated by measuring the time taken for the level of the internal liquid coolant in the converging chamber to drop from a predetermined level to a threshold level set by a single level sensor. The coolant drop rate of the internal liquid coolant in the converging chamber can also be calculated by measuring the time taken for the level of the internal liquid coolant to trigger two or more level sensors in the converging chamber.

[0073] Figure 12 A block diagram of a computer 700 that can be used in conjunction with embodiments of the present invention is shown. The computer 700 can be used as a control server or other computers described herein. The computer 700 may have fewer or more components to meet the needs of a particular application. The computer 700 may include one or more processors 701, one or more user input devices 702 (e.g., keyboard, mouse), one or more data storage devices 703 (e.g., hard drive, optical disk, solid-state drive), a display screen 704 (e.g., liquid crystal display, flat panel monitor), one or more accelerometers 705 (e.g., graphics processing unit (GPU), neural processing unit (NPU)), a computer network interface 706 (e.g., network adapter, modem), and main memory 707 (e.g., random access memory). The computer 700 may have one or more buses 708 coupling its various components. The computer network interface 706 may be coupled to a computer network 709.

[0074] Computer 700 is a specific machine programmed with one or more software modules 710 including instructions that are non-transitory stored in main memory 707 for execution by at least one processor 701, causing computer 700 to perform corresponding programmed steps. The article of manufacture may be embodied in a computer-readable storage medium containing instructions that, when executed by at least one processor 701, cause computer 700 to operate to perform the functions of one or more software modules 710. In one embodiment, software module 710 includes instructions for performing server management software or other software for leak detection and intervention disclosed herein.

[0075] While specific embodiments of the invention have been provided, it should be understood that these embodiments are for illustrative purposes and not for limitation. Many additional embodiments will become apparent to those skilled in the art upon reading the to which this disclosure pertains.

Claims

1. A cooling system for multiple servers, the cooling system comprising: Multiple cold plates are attached to the corresponding processors of the multiple servers; A coolant reservoir containing internal liquid coolant; and A coolant distribution manifold through which internal liquid coolant flows from the coolant reservoir to the plurality of cold plates. The coolant distribution manifold has a tapered chamber, a first level sensor for monitoring the level of the internal liquid coolant in the tapered chamber, and an exhaust valve connected to the tapered chamber.

2. The cooling system of claim 1, wherein the leakage of the internal liquid coolant is detected by triggering the first liquid level sensor based at least on the liquid level of the internal liquid coolant in the tapered chamber.

3. The cooling system of claim 1, wherein the coolant distribution manifold further includes a second level sensor for monitoring the level of the internal liquid coolant in the tapered chamber, and triggers the first level sensor and the second level sensor to detect leakage of the internal liquid coolant at least based on the level of the internal liquid coolant in the tapered chamber.

4. The cooling system of claim 3, wherein the triggering of the first liquid level sensor and the second liquid level sensor indicates that the rate of decrease of the internal liquid coolant exceeds a threshold rate indicating coolant leakage.

5. The cooling system according to claim 1, wherein the tapered chamber forms a conical funnel.

6. The cooling system of claim 1, wherein the tapered chamber is located in the level sensing section of the coolant distribution manifold, and the level sensing section is connected to the main section of the coolant distribution manifold.

7. The cooling system of claim 6, wherein the liquid level sensing section is threadedly connected to the main section.

8. The cooling system of claim 6, wherein the liquid level sensing section is connected to the main section via a piping system.

9. The cooling system of claim 8, wherein the piping system comprises hoses.

10. The cooling system of claim 1, wherein the coolant distribution manifold is vertically mounted in a rack housing the plurality of servers, the tapered chamber is located in a level sensing section of the coolant distribution manifold, and the vent valve is disposed on top of the level sensing section.

11. A method for detecting coolant leakage, the method comprising: Attach the cold plate to the processors of multiple servers; The internal liquid coolant flows from the coolant reservoir through the coolant distribution manifold across the cold plate; Monitor the liquid level of the internal liquid coolant in the tapered chamber of the coolant distribution manifold; In response to the liquid level of the internal liquid coolant in the tapered chamber triggering one or more level sensors, leakage of the internal liquid coolant is detected; and Intervention is initiated in response to the detection of a leak in the internal liquid coolant to protect the plurality of servers.

12. The method of claim 11, wherein detecting the leakage of the internal liquid coolant comprises: The first level sensor is triggered in response to the liquid level of the internal liquid coolant dropping below a first threshold. A second level sensor is triggered in response to the liquid level of the internal liquid coolant dropping below a second threshold, the second threshold being lower than the first threshold; The rate of descent of the internal liquid coolant in the tapered chamber is calculated based on the time taken to trigger the first and second liquid level sensors; and The leakage of the internal liquid coolant is detected in response to the rate of decrease exceeding a threshold rate.

13. The method of claim 11, wherein the intervention comprises normally shutting down the plurality of servers.

14. The method of claim 11, wherein the intervention comprises immediately cutting off power to the plurality of servers.

15. The method of claim 11, wherein the tapered chamber is connected to an exhaust valve.

Images